Meta-analyses suggest that aggregating children with externalizing behavior problems into groups has the potential of reducing the effectiveness of preventive interventions (Lipsey, 2006). There are two major gaps in the literature that limit our understanding of this phenomenon. First, intervention content is often confounded with the intervention format, so it is difficult to clearly disentangle the two issues about intervention effects. Few studies have randomly assigned children to a single intervention delivered in either group vs individual format. Second, rarely do intervention studies carefully study individual responses to an intervention vis a vis direct observation of group sessions, to identifiy specific processes and events that may enhance or detract from the intervention's effectiveness (Dishion &Dodge, 2006). This research proposes to do both within the context of a randomized trial. This research will be most productive if it can test variations in format for an intervention that has already established an evidence base and has the potential for wide dissemination. The current application will use the child component of the Coping Power program, based on a contextual social- cognitive model, as the focal intervention. The Coping Power prevention program has been demonstrated to have preventive effects on youths'aggression, delinquency and substance use in prior efficacy and effectiveness (Lochman &Wells, 2002a, 2002b, 2003, 2004) studies. The planned study will randomly assign aggressive children to one of two versions of the Coping Power child component. The Coping Power child component will either be delivered in the usual small group format (Group Coping Power: GCP) or in a newly-developed individual format (Individual Coping Power: ICP). It is hypothesized that ICP will produce greater reductions in substance use, children's externalizing behavior problems and delinquency at a one-year follow-up assessment, in comparison to GCP. This study will also determine whether children's characteristics (e.g., effortful control;outcome expectations;severity of aggression;callous-unemotional traits;impulsive decision-making;peer rejection and victimization;demographic variables) and group leaders'behaviors will influence children's susceptibility to deviant peer effects in groups in the GCP condition (based on coding of videotaped group sessons and leader ratings), and weaken the intervention's effects on substance use, delinquency and externalizing behavior problems. PUBLIC HEALTH RELEVANCE: A significant central public health issue involves the provision of evidence-based preventive interventions to reduce youth substance abuse and violence. Within the youth prevention field, a critical concern has arisen that working with at-risk antisocial children in group formats has the potential for deviant peer effects, whereby children escalate or maintain their behavior problems, rather than reducing them. This application offers an innovative opportunity to empirically examine conditions under which deviant peer processes are promoted or prevented during the course of intervention in a randomized trial, and will be the first study to directly compare the effects of group versus individual administration of the same evidence-based preventive intervention on long-term youth outcomes.